Inspiration
We were inspired by the large time delays between when citizens report an issue and when first responders are able to process the information and take action. We realized that these delays are especially pronounced in urban areas where there may be high volumes of calls due to the condensed population. We recognized that the speed of first responder action is crucial in building any community and set out to build a solution.
What it does
UrbanEye allows residents to call any phone number and explain any and all details of their emergency to a system that will present this information to first responders in the form of a summary that is organized in our user dashboard. UrbanEye is intended to be used when call traffic limits the amount of human attention each problem gets
How we built it
This project was primarily built using Node.js, with Twilio integrated to enable call answering and voice-based interactions. To generate responses, ChatGPT and its APIs were utilized, employing prompt engineering to guide ChatGPT's role and conversational position. Ngrok facilitated communication between the website, private backend, and public Twilio accounts, ensuring smooth integration between local and internet-based functions. Additionally, the Hugging Face API was employed for supplementary tasks like using All-MiniLM-16-v2 for basic sorting.
Challenges we ran into
The primary obstacle I encountered while developing 911 Responder was effectively integrating the diverse API services. Given the backend's critical role in supplying data for significant portions of the frontend through numerous API calls, this challenge was particularly acute. To complete the project, I had to rapidly adapt and familiarize myself with several new APIs. The most significant hurdle was integrating Ngrok, which was not initially part of my design but became essential as the backend matured. Learning Ngrok's applications quickly was a defining experience of the hackathon, highlighting the importance of adaptability and problem-solving in such a fast-paced environment.
Accomplishments that we're proud of
Using Ngrok could slow down our 911 Responder app, especially when handling lots of data. We also had to work hard to make sure the app didn't crash if something went wrong and to give users helpful error messages.
What we learned
Through building the project we learned the invaluable importance of a development schedule, particularly when tackling complex projects. By meticulously estimating the time and effort required for each feature, we were able to effectively prioritize tasks, allocate resources, and anticipate potential challenges. This strategic approach enabled me to overcome obstacles more efficiently and ultimately deliver the project within the desired timeframe. The experience reinforced the critical role of planning and organization in successful project management.
What's next for UrbanEye
Next up for UrbanEye, I will be finalizing our offerings for all kinds of calls. This will expand the project's utility and provide immediate development opportunities. Additionally, we will further explore the ChatGPT Whisper and Hugging Face APIs. The former was introduced by a mentor, and the latter offers features that can enhance the application. To ensure the highest level of security, we will also prioritize implementing robust data protection measures throughout the development process.
Log in or sign up for Devpost to join the conversation.